Biological Authority in Hybrid Computing: Reversed Control Hierarchies, Computing-Grade Neurons, and a Shared-Control Protocol with Simulation Results

Every hybrid bio-silicon computing system built to date treats silicon as the controller and biological neural tissue as a subordinate co-processor. This paper proposes reversing that hierarchy: building systems where living biological tissue holds control authority setting goals, making adaptive decisions, and directing silicon execution resources while silicon serves as the interface, translation, and I/O layer. We identify neuron longevity as the primary engineering bottleneck blocking all paths to practical biological computing and propose a genetic engineering approach to create computing-grade neuronsiPSC-derived lines modi ed with oncogenic and extremophile genes that are unsafe in clinical contexts but carry zero cancer risk on a chip. We present a three-con guration progression from silicon control (Con g A) through shared control (Con g B) to full biological authority (Con g C), and report simulation results from a 500-neuron spiking neural network proxy implementing the Con g B shared-control protocol. The simulation demonstrates Free Energy Principle convergence, adaptive silicon proposal re nement under biological veto, and structured sparse ring consistent with cortical dynamics. We establish an ethical framework requiring continuous welfare monitoring and mandatory termination protocols for biological substrates exhibiting indicators of unresolved distress.